Mario Odyssey Review

Mario Odyssey is 3d Mario’s return to the sandbox style of scavenger-hunt play featured in Mario 64 and Mario Sunshine. It features many hundreds of collectibles, called Moons, to collect through its levels (called Kingdoms), allowing you to progress from each level to the next after collecting a certain number of that level’s moons.

There are only 3 functional buttons, jump, hat, crouch. They can perform a range of actions based on context. Like crouch does a ground pound in the air, and then pressing hat throw will dive or roll depending if you’re in the air or ground respectively. Crouch and jump will perform a backflip on the ground. It’s interesting that they were able to condense the functionality of the game this much, but it’s kind of a shame that they didn’t use their excess buttons to more directly access functions like diving, even if it feels nice to groundpound into a dive. It can lead to pressing Y too soon after ZR, causing no dive to occur.

The sideflip requires you to have momentum before you can enter into the turnaround animation which cancels to the sideflip jump, unlike mario 64 and mario sunshine, making sideflips harder to do and more time consuming. Similarly, the spin jump requires a lot more rotation to initiate the spinning animation which transitions to the spin jump, and the spinning hat throw. Also, walljumping doesn’t give you much control over the angle you jump off, and doesn’t go as far as a regular jump, limiting its usefulness. Also unlike Sunshine, you can’t dive out of spin jumps, because dive is no longer its own button, but a function of ground pound. The crouch button will cause you to fall faster during a spin jump instead. Because of the nerfs in these moves, they’re much more limited in their functionality than their equivalents in Mario 64 or Mario Sunshine, which is kind of disappointing.

Every movement move has a niche, almost. Rolling is the fastest method of movement, especially downhill. Triple jumping gives you a high jump and works well with steps. Ground pound jump is the highest normal jump method. Long jumping lets you jump really far, but not so high (it’s technically outclassed by ground pound hat jumps for distance, but it’s faster to start up). Back flipping can be a bit faster to set up than a ground pound jump, even if it doesn’t go as high. Diving can cut your jump short, and both it and long jumping can lead into a roll. Spin jump has you hover for a long time, but otherwise doesn’t seem very useful unfortunately, especially since it takes so long to start. The side flip would be useful as a less effective and faster ground pound jump if it were easier to start. There is no fall damage, but there’s a stun for falling from a high height. I think this is an appropriate penalty. I don’t think fall damage would have added a lot to this game, and it’s a bit more forgiving for beginners.

Learning the full distance hat jumping tricks, both to jump as far as possible, and to climb as high as possible, is both fairly tricky and mildly technical. There’s a lot of states you pass through, and a lot going on in the process of the move. First you jump, then you throw your hat with Y, which can set mario’s rotation directly to the direction the stick is pointing, allowing him to make hairpin turns, then you press ZR, which triggers a ground pound, because you’re in the air, but if you tap Y during that ground pound, you’ll do a dive instead, and you continue holding Y, which keeps your hat spinning in place, so when mario touches the hat, he jumps off it. Additionally, in the dive state, you’ll bonk your head if you run into a wall, but after jumping off the hat, you’ll be in a neutral jumping state instead, allowing you to jump off walls, or even throw your hat again, but critically, it remembers if you’ve bounced off your hat already since the last time you touched the ground, and will not allow you to do it again.

The full distance hat jump and its components are amazingly versatile. The jump off cappy resets your air state so you can walljump, making this trick one of the best ways to scale walls, and to jump across long gaps. Since tossing cappy lets you rotate in a direction instantly, it can be used in midair for up to two hairpin turns. The full hat jump somewhat technical to pull off, and honestly took me more than a bit of practice to figure out how to do and eventually become comfortable with, though it was really easy to perform whenever I wanted after I got the inputs figured out. It’s obviously not the most technical platforming trick I’ve ever done, coming from Mirror’s Edge, but it’s really interesting that they crafted this series of mildly difficult moves on purpose, giving them the properties they have. This definitely exists as pretty much a archetypical example of my idea about including lower affordance tricks that enable higher skilled players to have more fun with the game. The game isn’t built to require this trick at any point to progress, even the very last few hidden stages, but it is built to enable its use pretty much anywhere, meaning the game plays differently as you progress from lower to higher skill.

Underwater controls are sort of like normal jumping controls, dunno if this has been the standard for mario games for a while, but it’s very different from mario 64, and frankly a rational choice. Crouch can be used to ground pound in the water to dive, and pressing hat throw after that will have mario swim forwards quickly, which is sensible. Cheep Cheeps even use a similar scheme, letting you swim directly upwards or downwards with a button press, instead of rotating freely and having a “gas pedal” like mario 64 swimming controls.

Cappy is a rather flexible tool for attacking enemies. He’s a mid-ranged attack that can be held in place to act as a wall between enemies and you, and he’ll hit things on the way back too. The spin toss and homing attacks add further versatility, as does his ability to grab pickups that aren’t moons. The hat throw notably has you instantly turn in the direction you point on the analog stick, ignoring normal rotational movement rules, which is part of what makes it so versatile compared to the dive by itself. It also hits the area immediately around you, and functions as an air-stall, much like the spin attack from mario galaxy. It’s nice how they were able to roll together the functionality of the mario galaxy spin attack into cappy while also adding new functionality, though it makes me kind of sad that he probably won’t be there in the next 3d mario.

Levels have a very mild interconnectedness within them, though usually have a straightforward path of progression on your first trip through. The horseshoe level design pattern is used a lot, with areas divided either by gaps or walls that can be crossed or scaled with advanced hat jumps, so there’s a lot of small sequence breaks all over the place. Levels are slightly bigger than 64 levels, I think. Except for lake kingdom, which is clearly smaller than most 64 levels. There’s linear sections inside hidden areas, like there were in sunshine, but they have less “density” of level elements than sunshine levels, and consequently aren’t as fun/challenging as say, Noki Bay 6. I do however like that these linear challenges almost always have a second, more challenging, and usually slightly secret moon to collect, even if it’s still not very challenging. Levels in general are not hard to complete. The only significantly difficult level is Darker Side of the Moon, and that’s only because there’s literally no checkpoint for the entire thing. I’m not really impressed by the difficulty of even the harder bonus stages, like the hatless bullet bill stage on dark side of the moon, which I was able to beat in less than 5 tries.

I think there are way too many moons frankly. Not a fan of the collectathon elements in general, though it’s more tolerable here than in other games. You basically have a world with a ton of moons, you have a goal of a smaller number of moons to complete, so you can get by with only picking up a few. Progression up to the first ending on the moon has a pretty steady pace, because the number of moons in each level is so dense, so you’re constantly tripping over moons and get into and out of levels fairly quickly, without much down-time in hunting out the last few moons. The later phases of the game, like dark side of the moon, and darker side of the moon, require you to hunt down a ton of moons before they become available, and this is where it gets tedious, since as you collect more moons, you leave behind the more obscure ones, and the remaining moons are less densely packed together, making it more time consuming to find and reach each remaining moon. Which leads me to say, there’s way the fuck more moons than there have any right to be. A lot of the moons are totally trivial, just find the thing that’s out of place. Along the main progression path of each level it’s pretty okay, because you have more or less structured content that you can follow that’ll give you close to enough moons to advance for solving that world’s issue, fighting the broodal + boss, and you can pick up additional hidden moons as you go. The ability to buy moons at the end of the game is weird. I guess it’s there because it gives you a way to cash in your coins and get to the 250 and 500 moon goals more easily, since the scavenger hunt gets really protracted and tedious at this point in the game. They also have the clever move of adding a moon rock to each stage, which spreads more moons throughout the stage once you’ve finished the moon level with bowser, which can help make the distribution of moons more dense at this late stage in the game. Getting moons as achievements from Toadette is a complete pain however, you need to sit through dialogue and the animation each and every time. This was not a wise inclusion in my opinion, especially since most of the achievements are extremely banal, like collecting a large number of coins, jumping a lot, throwing cappy a lot, etc.

The camera is REALLY GOOD. The max speed is maybe a bit too slow. A ton of areas have meticulously placed camera hints, or will lock the camera at simple angles for framing the action if that area has a mostly 2d layout. Most of the time they allow you to adjust the camera afterwards and will keep it at your adjusted angle rather than defaulting to the hint angle until you exit and reenter the camera hinted area. Some minor flaws are it can still be easy to get large objects in the foreground obscuring the camera (I’m used to moving it manually, so I didn’t notice much, but my dad had it happen a lot) and that it can obscure where mario is in tight spaces. These are fairly standard problems, and it’s hard to avoid the tight spaces problem without causing MGR style camera issues, which are arguably worse.

Each of the broodal bosses is deliberately designed with a quick way to kill them, which is pretty cool. Like, they each function in cycles, where you normally need to wait through their attack and then they set up for you to get a guaranteed attack on them, but there’s always a way to interrupt their attack cycles, like with the purple one, you can hit her bombs back at her with a good angle to knock her hat off early, and during her UFO phase, you can hit the bombs she drops up at her. The blue, yellow, and green ones, you can stomp on them when they turn into hats, forcing them back to their main phase. During the wooden robot boss fight, you can even get back on its head with a fancy extended hat jump. Notably, the Broodals are not invincible during phases where the robot is not knocked down, so they can be attacked when the robot is upright, which is an uncharacteristic decision by Nintendo. Inclusions like this and the existence of the extended hat jump at all indicate that Nintendo is getting a better idea of how to deliberately cater to speedrunners, without compromising on core gameplay.

Levels have a lot of horseshoes, places where goals are placed near the start, usually separated by height, with little footholds, so you can get from the start to the end (or start to the middle, middle to the end) by doing the advanced hat jumps. Really obvious one is in the lake kingdom, with just a high wall separating you from the end of the level. Many levels have a clear (winding) path of progression you’re supposed to follow through the level and usually some side routes or back routes through the level that are shorter, or give different access to the level (cascade kingdom, sand kingdom, wooded kingdom, luncheon kingdom). Others are more sprawling, having a big open area with multiple objectives or no primary objective (new donk city, seaside kingdom, lost kingdom, mushroom kingdom) A couple are out-and-out linear (cloud kingdom, ruined kingdom, bowser’s kingdom, moon kingdom, dark side, darker side). Most of the winding path style kingdoms open up into total freeroam after having the main quests dealt with, thanks to new shortcuts opening up. What I would have liked to see would be more of a happy medium between the more linear challenges, and the complete freeroam areas. I would have liked to see the power moons consolidated more into more worthwhile challenges instead of a lot of, an obscene amount of, scavenger hunt shit. Moving from kingdom to kingdom works pretty well, as you only need a few and you have the freedom to improvise and pick stuff up as you go, but in the later stages of the game, such as unlocking dark side or darker side. It’s tedious to need to comb for almost every single moon in an area. Speedruns of all moons are 10 hours long. Darker side of the moon is just shy of 4 hours. To me, in the context of this game, this is an indication of a lot of filler content. The achievements with Toadette are in particular, a major time-waster to collect them all, and strike me as really unnecessary. Do we really need progression tied to throwing your cap X number of times or collecting X number of coins? It would be nice if every power moon were a challenge to collect in of itself, and if they preserved the way it’s interesting to route power moon collection across the level, because that is still cool. It’s nice to have the improvisational aspect of figuring out which power moons are easiest to grab in the most direct line as you zip across the level’s obstacles through weird routes, but with the compromise that if you’re not speedrunning, you don’t need to engage with any of that and many of the challenges are just kind of simple and one-note if done the intended way, which is kind of disappointing from a casual play perspective. Darker side of the moon is probably the hardest stage, but also the most linear and restrictive in how it can be tackled. It doesn’t emphasize the usual multi-threaded strengths of mario level design. It’s just a sort of tough execution challenge with no checkpoints at all. A lot of the linear bonus sections that mimic Sunshine’s “secret” stages are similar to this, and don’t have the same diversity of platforming options as even Sunshine’s secrets offered.

It was also cool how new donk city littered the city with cars that could be jumped off of, and poles that could be flicked to get a boost upwards or forwards. New Donk City has a lot of routes across it, and most of them involve platforming, which works really well with the different height, multi-terraced buildings. New Donk City is a stand-out for this style of sandbox level design. It’s a shame it wasn’t larger, and maybe that it didn’t have more girders going between buildings.

It’s disappointing they didn’t do more with the moon physics on the moon level. You have an outdoor platforming section which is really small and easy, even by the game’s standards up to this point, and all the indoor sections have normal gravity. You fight a boss rush with the Broodals in low-G on the dark side of the moon, but that’s still a disappointing use of the physics changes. It’s interesting how it’s possible to turn so much more in the air in low-G, even off a walljump, but there isn’t any level design to take advantage of this, so again, it’s just kind of a waste.

The koopa races are cool, and are always set in a part of the level with a good linear goal, but which is also mutli-threaded. Or they start at the beginning of the level’s intended progression, with the goal at the end of that progression, which demonstrates how those levels are multi-threaded in of themselves. I’ve seen a number of different ways of beating each of these, which suggests the depth of the game mechanics and the level design.

I actually had my Dad play the game a little. He doesn’t really play video games, but he was interested in this one when I got it, remarking that he wanted to give it a try, having played the old ones, but then the new ones got too complicated for him and he lost interest. Given the game essentially operates everything on 3 buttons, I decided it was worth a try, to see how hard it would be for him, also as a case study of someone learning 3d action games for the first time.

To give some context, he was completely unfamiliar with the controller, didn’t know what any of the buttons did. So I had to show him that the A button was confirm, and B was cancel. As he played, he frequently looked down at his controller to double check the buttons he was pressing.

He was used to using the R button to lock the camera behind him, probably from Tomb Raider 2 and 3 having a similar functionality forever ago, but was not used to using the right stick to point the camera. When he wanted to see something offscreen, he’d frequently either try to move mario over to get it on-screen, or strain his head trying to look offscreen. He got REALLY disoriented when he tried the Jaxi for the first time for example

His movements with mario on the control stick were very jerky, reminding me a lot of when I was first learning to drive a car, making microadjustments, rather than one continuous fluid motion. Over time he learned to move more naturally and slowly got used to using the right stick to aim the camera, forgetting to use R to center it now that he had a new tool instead of smoothly integrating both. He also got confused about which direction to hold the stick to rotate the camera in the direction he wanted, and very frequently ended up with the camera at a much higher or lower angle than he wanted. Also he’d frequently get foreground objects obscuring mario, but because he wasn’t used to using the camera, he’d end up not moving the camera to get it out of the way automatically. Or in tight spaces, mario would be obscured, and he wouldn’t be able to tell where he was.

He had trouble avoiding a number of things, even in the 2d sections and reacting to things coming at them, often overreacting, and walking straight into obstacles, or jumping into bottomless pits. He got a bit frustrated by losing, even though the penalty wasn’t that high, but I could definitely see him improving over time, even if he didn’t notice it himself. At first he wanted to collect all the coins he saw, but I pointed out that he had like 500-600 coins already and he didn’t need the coins that badly. He wasn’t used to building a 3d model of the area from what he saw and frequently got turned around while navigating, going past where he wanted to be, or moving out of place and not realizing where he was anymore.

He also tended to miss details, like all the buildings in the town in the sand kingdom, walking straight through without paying attention to what was around him, though it may have been possible he was just really focused on the objective. I mostly gave him light hints and reminders of the controls to help him along, but generally tried to allow him to fail for himself instead of telling him the answers to problems. I did point out a few hidden moons however.

Versus the purple broodal, he frequently walked into her bombs instead of avoiding them. His ability to aim the analog stick for throwing cappy was also not great, doesn’t seem to connect the direction of the stick to the character on-screen very well yet, though judging perspective can honestly be tricky. After he failed at the broodal fight multiple times, I ended up doing the fight myself, to show him the advanced tricks for the fight, and beat her in one go, stopping before the last hit and killing myself so he could give it a shot. He was then able to get her on his next try.

The difficulty being as low as it is makes it a good fit for someone who has basically never played a game before, which I think was the intent, but it’s pretty obvious that it’s still really difficult for someone to learn how to use a controller, remember the buttons, and aim the camera for the first time if they don’t have a background in it, even if most experienced players find that mundane. It’s probably worth remembering that players like this exist, but it’s hard to say what should be done to help them out. Mario Odyssey happens to be really convenient for this purpose, but Dark Souls or maybe even Zelda would be insurmountable in comparison.

In this way, Odyssey’s mildly multi-threaded level design that allows for advanced players to challenge themselves, and gives them harder moves to figure out and use is a decent compromise. I don’t think the scavenger hunt style design really suits the purpose of benefiting the beginner or expert however. Beginners get lost easily, and experts are likely more annoyed at having to comb the levels.


So, the more interesting captures in the game I’d say are the cheep cheep, uproot, hammer/fire/fryingpan bros, tropical wiggler, shiverian racer, gushen, lava bubbles, Pokios, bowser, and yoshi.

Overall my remark on the captures is they’re mostly pretty one-note and not that interesting, which I honestly should have expected since before the game came out. Trying to make a ton of things deep usually doesn’t work as well as adding more depth to a few things (better to go multiplicative instead of additive). All the captures suffer in that they can only use 2 buttons, B/A and Y/X, since ZL/ZR are used to release the capture. So all the captures are pretty simple, and I’d say the ones that are most interesting are the ones that do something unique with their control scheme, despite their limited button count.

Cheep Cheep let you move freely up and down underwater, which I think is fairly cool in a Zone of the Enders kind of way. Can also spin to go faster and attack. If directed out of the water, they can jump higher out of the water than mario can alone, giving them a unique utility in that regard.

Uproots are pretty cool. Instead of jumping, they grow taller. Then, when the button is released, they snap their legs up to the top, and do an itty bitty jump. This means they can touch anything up their entire length while grown up, and can walk off ledges and jump up overhangs. It also means you need to be careful of their feet when grown up, which is an interesting consideration to have which jumping normally does not.

The hammer bros are neat mostly because their standard form of movement is jumping, and you can jump out of their jump, giving them a sort of double jump. Also they hurl their respective weapon in a random arc in the general direction their facing, which can be kind of annoying when it doesn’t cooperate, but it feels kind of cool to use.

Tropical wiggler are unique, extending themselves across gaps, carrying their length along like it’s a flexible rope. Similar to the uproots, they’re vulnerable along their whole length, and their sections are usually designed around this, creating a unique “platforming” challenge, like those old games where you need to draw a line across an area without letting the line get hit until it’s done.

The shiverian racer can be kind of interesting. To go fast, you gotta bound every time you hit the ground, and keep steady control over which way you’re going, as well as the angle of the ground you land on, which can be an interesting combined challenge.

Gushen are mildly cool. They can jet along the water extremely fast, but also adjust their height in the air, practically flying above the water, much like a cheep cheep does underwater, except to descend they need to drop like a stone. They can also hit things by either getting above them, or pointing away from them, both of which can be tricky to manage, since you’re affecting your movement at the same time as you do this.

Lava bubbles are kinda basic, their big deal is that they can jump really high, and only move in lava, so their levels tend to be about jumping between puddles of lava. They don’t have much air control either, so you need to build up speed to get a good jump with them, and you generally have a lot of commitment. It’s also cool to aim for the tomatoes to make new puddles to jump into. The cookatiel fight with the lava bubble is really neat, platforming onto blobs of lava suspended in the air.

Pokios are really cool. They’re kind of a play on the poles and forks which were featured in the game before them, but they’re much cooler. They can attack with their spear beaks, but also poke into any soft wall and flick themselves in any direction up or along the wall, giving them very unique platforming challenges. Their moveset is deep enough and has enough potential that you could probably build a whole game around them and it would be pretty decent. To use them well, you need to carefully time when you poke into the wall, you need to flick accurately and move in the air to get around corners, while still orienting yourself towards the wall so you can peck into it. Using them effectively is a real challenge, and they get a few really nicely designed sections. It also helps that they can redirect bowser’s bombs like billiards balls and do a spin attack with their spear beak, also a motion control air stall.

Bowser is basically Strider. He can attack while running, triple jump, and shoot fireballs. Honestly, on review, his section is better mostly because it has some neat level design, rather than because the capture is particularly interesting, though the capture does go hand-in-hand with the levels, placing blocks in your way that need to be attacked to get past. Also the falling boulders, ground falling out on you, and rolling rocks help make it interesting. There’s a lot going on all at once, and you’re expected to keep moving while dealing with it.

Yoshi really suffers from the lack of a groundpound button, since he’s basically just Mario with a tongue and flutter kick. The tongue can be used to grab stuff like enemies and fruit, or to grab onto walls and hang onto them, then awkwardly dive off the walls. Flutter kick lets you hover, which is pretty okay. It does not allow you to gain any height however. Yeah, Yoshi’s kind of boring.

Overall I think the most appropriate rating is 7/10. It has the spark with a neat moveset, but it’s not an amazing game. I’d say most of the game’s faults come down to not pushing the challenge and multi-threading in its level design as much as it could. I’d like to see a game with more routes through its levels, and all of those routes being more challenging, even if the challenging content is reserved for the end. Still, it represents a small shift in the way Nintendo is thinking about their games lately, trying to be more accommodating of multiple skill levels.

Critique of Super Smash Brothers Melee Review and Analysis

The slowness of the switch between Zelda and Sheik is not an engine limitation. Both characters have their files loaded when the game begins, so that they’re both in memory and switching happens as fast or slow as the actual animation. This is not true in Brawl however, where the other character is loaded on the spot.

The example of mewtwo’s up throw killing captain falcon sooner than fox is a bad one, because throws do not differ their level of knockback based on character weight. Weight only affects the length of the throw animation. For mewtwo’s up throw, the only character-specific factor that affects how far the character goes is their gravity, not their weight. A more accurate example would have been a move like fox’s up smash, where both weight and gravity can affect it.

The speculation about Sakurai’s intentions with Peach is really bad form for a review/analysis. Unless you have a statement from the developer, it’s best not to guess what their intentions were, unless it’s significantly more obvious than, “Sakurai probably thought girls were bad at the game, so he threw them a bone by making a powerful but easy character.” It’s a lot easier to guess about intentions based on changes between games, than something like expecting a certain response from a particular audience.

I think the further speculation about Bowser is even more unwarranted and unsubstantiated. My general advice for someone trying to do video game analysis is, stop caring so much what the director was probably trying to do, analyze the thing as it is. The author is dead, all that exists is their work. We can analyze the work itself and determine what it is or isn’t, but it’s nearly impossible to guess why it’s that way without direct statements from the developer. Further, it’s counterproductive to let the “why” of a work dominate analysis of it, because the only thing that makes the work is the work itself, not the developer’s statements on it. The work will be what it is, regardless of what the developer says it was supposed to be. To understand the value of the work, to understand its implications in comparison to other works, to create a basis for us to create or appreciate new works, we must be looking only at the results, not the intention. If you get too focused on the intention (which you don’t know definitively in the first place), then it distracts you from the results.

Something can completely contradict the intention of the author, but be good, something can completely accord with the intention of the author and be bad. The author’s intention doesn’t matter, only what’s left when they’re done.

You forgot that one of Roy’s other significant advantages is his down tilt, which pops enemies up, instead of outwards. This makes it less useful as a ledgeguard, but significantly helps Roy’s combo game versus Marth (even if Roy’s combo game is still overall worse).

Okay, you’re doing the author intention speculation thing again, “Sakurai clearly doesn’t find the simpler stages as interesting, seeing as battlefield and final destination are some of the last you unlock” Super amateur. This is completely baseless. Mushroom Kingdom 2 is frequently one of the last stages obtained by people trying to clear a new copy of melee because obtaining the birdo trophy is so rare. Does this mean Sakurai hated Super Mario Bros 2?

If you’re going to engage in post-hoc rationalization like this, couldn’t I also say that he probably valued simple stages the highest, by making players work hardest to earn them?

C’mon dude.

C-stick can’t SDI, only ASDI. The held direction of the C-stick will override the control stick for the exclusive purpose of ASDI. It cannot make your character move during hitlag, only the control stick can.

If you said Sakurai 90% less, then this video would be 90% better. I review a lot of video analysis, so I was excited to see one about my favorite game, but as of 15 minutes in, you’re not trying.

The premise of the video, “Examining the decisions made by the developer” is flawed.

You don’t know what decisions he made, you only have the results. There’s tons of interviews by him, you can find them over on I’ve read nearly all of them. You’re not citing anything he’s ever actually said.

Who cares whether adventure mode took much time to make? Analyze whether the mode is fun. What it does or doesn’t do successfully. Is the level design of the stages good? Are the encounters good? How do they randomly or deterministically vary between sessions and character picks?

I’m getting more and more frustrated as I get further into this video. If the whole thing is just guessing what sakurai wanted to do, then this isn’t a very useful review or analysis.

You use the term, “Momentum preservation” twice without explaining what it means, just saying it makes the game feel good. Of course I know what it means, but you can’t assume that everyone who views this video does. It would be more clear to say, “transferring ground momentum to air momentum” rather than “momentum preservation” like it’s a key word. Given you don’t explain this clearly, and just play clips, it’s hard to tell what exactly you mean by this.

Also, WHAT. Street Fighter 5 does NOT have an 8 frame buffer. Are you insane!? It has a 2 frame buffer. If you want a game with an 8 frame buffer, you should go for Brawl or Smash 4, which have 10 frame buffers each. Dude, you had a game with an excessive buffer sitting RIGHT THERE next to Melee, and you decide to go cross-franchise, cross sub-genre, to a game that barely demonstrates what you’re trying to claim. Are you getting mixed up by the 8 frames of input delay meme? Because Melee has 4-5 frames of input delay (the extra frame of delay oscillates over time).

AGH, the shorten window on Fox and Falco illusion isn’t 1 frame, it’s 5 frames! Fact check your shit please! It’s not that hard to google fox and falco framedata.

I do think for once that you have a point on the developer speculation here however, there are a number of intentionally implemented mechanics (ones that are specifically coded rather than being emergent) with very small frame windows, which are very difficult to notice, such as shortens, fast falls, rest, L cancel, V cancel, light shield density, power shield reflecting, etc, and this suggests that sakurai wanted to implement things that only more dedicated players could master.

Then you go straight into more baseless speculation when it come to why DI was put in the game. Come on. Also, you failed to mention that hitstun was reduced in Melee compared to Smash 64. And also, SDI existed in Smash 64 already, called PI (position influence) by the Smash 64 community. (I didn’t watch your 64 vid, so if you covered it there, I’m sorry).

Also, “more options at any one time than a traditional fighter”? Dude, you know how to count, right? A traditional fighter has 6 attack buttons. Melee has 2. A traditional fighter has 12 ground normals, smash has 7. A traditional fighter has 6 jump normals, smash has 5. A traditional fighter usually has at least 3 special moves per character, and different versions of each depending on the button you press. Every character in Street Fighter 2 has more moves than any smash character.

Smash’s DI system only works in Smash, because it’s based on juggles and angles of knockback. I’ve considered ways to implement DI into other fighting games, and it just doesn’t work based on the control and combo systems. At best it might work in a game like Hokuto No Ken, or Marvel 3, but combos are much tighter in other fighting games, including juggle combos.

Smash Bros makes up for having less attacks by having more detailed attacks, and allowing you to move as you perform attacks, and between the linking hits of combos.

It’s not a unique sped-up animation for successful L cancel, it’s the same animation, but sped up. This is trivial in any 3d game’s engine, even in those days.

Also, seriously, stop the speculation. It’s not helping your video any.

I like the perspective on bugs and glitches overall. Works for me.

I think it’s worth mentioning that glitches exist in tons of competitive games, Quake, Starcraft, Basketball. Any game with a ruleset that is robust enough has some unintended emergent effects of multiple mechanics.

Glitches weren’t banned because they were undesired for the competitive experience, they were banned on the basis of having no counterplay, things that allow for indefinitely stalling the game, or . One thing that the rulesmakers didn’t want to do was curate the game on the basis of what was desired or undesired. It was more about trying to keep the game fair and prevent strategies without a counter, things that count as an “I win” button, like the freeze glitch or luigi’s ladder.

Similar to wavedashing, the ability to cancel normals into specials in street fighter was actually the system working as intended for the most part. I’m explaining this more as a history lesson than a criticism. In SF2, there’s actually a 5 frame window at the start of every move that can be canceled into specials, probably so specials would be easier to trigger if you mashed the button and triggered a normal before you were done inputting the special move command. However, if you hit someone, there’s hitstop, which also freezes the window for this cancel, allowing you to cancel into the special after hitting your opponent (and accepting inputs for this cancel during the ENTIRE hitstop). So again, it’s stuff that was intended to happen, happening in a different way, and it became the basis of fighting games in general.

The other fact of the matter is, playing with only one third of the cast is pretty average for most fighting games. Most fighting games only have a small number of characters that are competitively viable. Well-Balanced fighting games are a recent trend and have not been the norm across the genre. And as you said, there’s a lot of different ways to play those top 8, so it’s not a big loss. Also, if you played without competitive rules, then the worse characters become even less viable. This is just an inevitability of getting good at a game, some characters fall off.

For reference, 3rd strike only has 3 viable characters, chun, yun, and ken. Super Turbo has arguably only one viable character, Old Sagat (kind of debatable, the matchup spread in super turbo is funky). CVS2 has a small number of viable teams, same for MVC2, KOF XIII, Most Tekken games, Most Mortal Kombat games, and pretty much every fighting game game made before 2009. Melee was made in 2001, long before the balance trend happened, and frankly, having a fun game is more important than balance.

The character summaries and their place in the competitive metagame is good. Good descriptions of what each character can do and why it puts them where they are.

I think it would be slightly more fair to say the wobble is why icies are viable, not their grab game.

This section was relaxing after the prior sections, a lot less baseless speculation, a lot more laying down facts, a lot less inaccuracies.

I also love the description of the 15 frame reactionary blindspot. I’ve been talking about this for a long time, but I’ve never really seen anyone else cover it except you and M2K. You illustrated it fairly well too.

Would be slightly better if you mentioned that reads are a thing players can actually do, people are designed to sync up with people via something called Mirror Neurons. This happens inconsistently, but sometimes you just get a really strong sense that the opponent is going to do a certain thing. Studies have shown the capability of mirror neurons to predict the actions of other people before, lighting up before someone else performs an action. We don’t have conscious access to the results of these however, they filter up through our subconscious.

Good job on this part.

A lot of games have arbitrary execution factors like L canceling. See starcraft, or shooters. There isn’t a risk/reward to always producing new worker units on time, you need to continuously do it the whole game, or you’re at a flat disadvantage. (okay, not the whole game, but until you reach a later stage of the game where you don’t need additional workers)

L canceling has a risk/reward in that if you press it at the wrong time, you’ll get a longer landing animation than the L canceled one. you need to read the situation and try to always nail the L cancel, and trying to go for one timing instead of another. The L cancel timings can vary significantly between whiff, hit, and hit shield, or hitting someone’s tilted shield. They can also vary with multihit moves like fox’s drill, or vary even more significantly on ice climbers, who you can hit twice before landing.

This is like fast getup in Street Fighter, you almost never want to intentionally miss your getup. It’s a matter of recognizing the situation and timing your input right.

I agree that removing it (and halving the duration of all landing lag) would probably be fine, but it’s worth understanding that it’s not just an arbitrary skill check, and that arbitrary skill checks aren’t inherently bad either.

“Small chance to miss a backdash randomly”

There isn’t a small chance to miss a backdash randomly. Math.Random() is not called when you try to backdash. I realize that backdashes have been compared to randomness by some players, but it’s not actually random, it’s a result of the controller polling in the middle of moving the stick to the back position, instead of once it’s all the way there. There’s nothing random about this, It’s just very hard to get the stick from the neutral position all the way to the backdash zone in one frame. Players like Druggedfox even argue against mods like UCF, saying that backdashing consistently is a matter of getting good at the game.

Even if your intent was to say that consistency on this technique is so poor as to be random, calling it random is not the right thing.

Simultaneous grabs need to be resolved somehow, and a lot of fighting games use player number to resolve edge cases like these. Super Turbo resolved same-frame grabs randomly. SFV does the same for same-frame command grabs (based on the framecount % 2). I agree that it’s pretty dumb that Melee does this based on port priority. I’d personally prefer a 1 or 2 frame throw tech window, where the throw release animations are played by both characters. This would help resolve simultaneous throws without giving either character an advantage, much like guilty gear.

(Lack of) Level Design in Stylish Action Games

What do you think of the level design in action game arenas? Specifically how most fights that occur in games like DMC/Bayo are pretty barren. I get that fights should be engaging regardless of the level but I don’t see why arenas like this aren’t more common.

I mean, those games just don’t work with level design very well. Have you played the platforming sections in DMC3 and 4? They’re pretty awful. Platforming in Bayo and MGR is pretty bad too. All these games have character motion that makes characters very difficult to line up on a specific spot in comparison to platformer games like mario. All these games have jumpsquat animations and jumps that lock you in specific directions and don’t give you much air control, if any. This works really well in combat, it means you need to commit and choose options relative to your opponents, and you barely notice it there because all your actions are deliberate and straightforward, but for navigating geometry, it’s hell. Have you tried playing a fighting game which didn’t design its jumps this way? Like Super Smash Flash 1, which used more conventional jumping controls? In Super Smash Flash 1, unlike real smash bros games, when you press jump, you jump instantly, and you can switch your facing direction at any time during your jump.

DMC3 actually had a number of arenas like that. Like, there was this one hallway with corkscrews on the walls. Another with multiple platforms of different height that rotated. DMC4 had fights with enemies on disappearing platforms (I think that one worked rather well actually).

At 5:38

At 2:48

I think that detailed level design actually works against these games’ best attributes, because of the way they chose to specialize in their character controls. Other games, like Dark Souls, have better integration with their level design, which they can afford because of the different, more close-up control style of those characters. It’s a bit tricky for me to distinguish exactly what attributes make it this way though.

I think arenas like the above from transformers devastation aren’t more common because coming up with subtle level design elements for games like these, that add a twist, but don’t ruin things, is hard. It’s like trying to make additional tournament legal stages for Smash Bros or maps for Starcraft. You gotta follow a certain template or things get gimmicky and frustrating. Of course there’s probably room for more experimentation in the future. These games have good enough combat systems to let the combat alone carry the game in the absence of level design, which generally tends to work out.

Making a Fair Doppelganger Fight

What sets Devil Hand apart from the other kinds of doppelganger fights? (i.e. Dante) Or does it just show how good God Hands moveset is that it can also work for a Boss?

Well here’s the trick, God Hand doesn’t use Gene’s actual moveset on Azel or the Double God Hand Gene at the end of the arena. All of their attacks are reactable, and unlike you, they have the ability to block like common enemies, and even have a few new attacks. When you use roulette wheel moves, they’re effectively instantaneous because time slows for the cinematics. When they use it, it’s all in real-time. Continue reading

Leveling the Playing Field for Unskilled Players

Is it possible or necessary to level the playing field in fighting games between skilled and not skilled players?

Nope to both.

The only way you can level the playing field is by making skill less influential on the outcome of the game. You can do this either by reducing how difficult the optimal outcome is to produce, or by reducing the amount of advantage conferred by the optimal outcome relative to the least optimal outcome. Or you can do this by having randomness be a much stronger determiner of who actually wins the game.

Of course, all of these things will piss off skilled players. You might have some limited measure of success depending on the implementation though. I’ve been arguing that randomized bullet spread has been bad for years and should be replaced with damage dropoff relative to distance to get the same effect and I don’t think anyone agrees with me on that.

Since the difference between skilled outcomes and random outcomes are difficult to distinguish, especially when the two are combined, randomness seems to help some games out, like Hearthstone, Call of Duty (Red Orchestra devs once called CoD out for this), maybe SF2 (never been explicitly called out for this, but it is the most mainstream popular SF game, so it’s possible the randomized damage and stun had a role), and Smash Bros with items on. Randomness mixed with skill tests allows weaker players to feel good about their skills when they occasionally win, possibly adding to the overall “stickiness” of the game for a wide audience, while still allowing players to build consistency overall.

However Luck can still be beneficial from a business perspective. This talk by one of the famous designers of Magic: The Gathering basically makes the argument, “Luck is really good at attracting players, introducing more luck into a game angers existing players, so you should start out with as much luck as you think the game needs, then reduce it over time as the playerbase matures.” And it presents the case of Team Fortress 2, which followed his pattern of high luck early on and reducing it as the game grew older, which was pretty successful for TF2.

Smash Bros Melee had a natural release valve in this way. The default modes and stages have items and other random effects, but the game offers options to turn these off, so as the game matured into its competitive format, players had the ability to turn off most of the randomness and get serious.

So to build a successful game, randomness helps to attract an audience, but it should probably have patches to reduce the effect of the random elements over time, options to mitigate the randomness that are made low-affordance on purpose so the community can slowly discover how to mitigate or remove randomness, or variant rules to remove the random elements.

The Pain of PC Fighting Games

What do you think of this? Should fighting games move to PC to get the sponsorship bucks and move out of the kiddie pool?

There’s an additional issue he didn’t address, the difficulty of setting up controllers on PC. Individual users don’t tend to notice it, because they’ve set up their PCs to run their controllers just fine, they’re not having a conga line of people over to each individually plug in their controller of choice and somehow make it work with the operating system and then with the game. Continue reading

Hard Games vs Journalists

Can a review still be trustworthy even if the reviewer is bad at the game they’re critiquing?

I’d say that someone’s skill at a game is directly related to their ability to comprehend the game. There are certain insights that will only reveal themselves to you as your skill at the game improves.

That said, it is still possible for someone who is bad at a game to develop these insights, it’s just unlikely. If someone is good at a game, that is a weak indicator that their insights will be good. If someone is bad at a game, I’d say that’s a strong indicator that their insights will be bad.

I take the position that reviews should be written well/descriptively enough that you do not need to trust the reviewer to agree with their conclusion. From this perspective, it should not matter if someone is good or bad at the game, their review should be descriptive enough to be useful regardless.

So basically, if someone is bad at a game, don’t expect much from them, but also don’t discount their words entirely. Listen for whether what they’re claiming seems plausible or implausible. Listen for whether their claims seem to be dependent on their level of skill. A good reviewer (but bad player) can theoretically extrapolate beyond their own level of skill to deliver accurate insight, I just think it’s unlikely.

In this case, this video is shameful. They originally titled the video, “Cuphead: It isn’t easy”, but have since retitled it to what it is now. The tutorial is especially sad, when the guy cannot figure out he needs to jump off the block and airdash, and spends 2 whole minutes running into a wall.

I don’t understand not just how someone can be so incompetent as this whole video, but also how someone like this would become a games journalist. I’d expect that people with a high enough interest in games to become a journalist would be capable of putting 2 and 2 together.

We keep seeing footage like this slip between the cracks, previously with Polygon and Doom. Expect to see less footage from journos now that this happened.

Looks like that Cuphead guy made a redemption video or something. Not sure if there’s much to be said about it, but any thoughts?

You can tell he replayed the tutorial level until he got a good take, because there’s no coin at the end. Still pretty poor coordination, and doesn’t really excuse that he didn’t get it the first time, because it’s the type of thing that the average person wouldn’t take nearly as long to get their first time if they had played practically any other video game before, or had a reasonably long history of playing games. Pretty sure Arino from GCCX wouldn’t struggle as much.

I’m kind of surprised the cuphead guys didn’t update the tutorial with a barrier that can only be passed with an airdash, without jumping (such as a gap with a ceiling above it to prevent you from jumping across), then a barrier that requires both jumping and airdashing at the same time.

You can tell his ability to switch from one button to another is still pretty bad, and he kind of goes on auto-pilot when moving and shooting at the same time, like he can only manage one thing at a time in his mind, but he’s also not playing panicky and running back and forth without purpose and running directly into obstacles anymore.This is still below the skill level of what I’d expect of a grown adult, much less someone who has covered games for 22 years. He seems like someone who has a hard time moving while dribbling a basketball.

I think George makes a good point in this blog post.

“The few “real journalists” I admire and follow in this business didn’t earn their reputations off of reviews, and the stories that made them big weren’t reviews. Ten years ago, Geoff Keighley made it big writing lengthy interview-driven feature pieces documenting the development of Half Life 2 and Portal 2 before they were even out. Four years ago, Danny O’ Dwyer was making flashy video essays boiling industry-trend criticism into common-person polemics on Gamespot. Last year, Laura Kate Dale released a string of infamous leaks revealing future Nintendo plans before the company could control the release of that same information. That’s what I consider “journalism,” and it has nothing to do with how good they are at games.”

There’s a lot to be reported about games that can be reported inerrantly without any sort of skill at games. You can report a lot of information that people want to know about games that has literally nothing to do with how games are played. And Dean Takahashi mostly does report that type of information. He has done roughly 2 reviews out of 14,000 non-review articles.Dean isn’t good at games, but for the vast majority of the content he writes, that doesn’t matter. He isn’t expressing opinions about how games are or the way games should be. He’s providing coverage about pretty much everything other than that, and that’s okay.

Cool Game-Related Anime

What do you think about game situation analyses in media like Hunter x Hunter and No Game No Life? I personally think they’re kind of cool, but do you think they misrepresent how games are actually played?

I ended up watching NGNL because my brother was really into it. On reflection, it didn’t really have anything to do with games, and kind of made up the rules of every game as it went. Continue reading

What Type of Future Tech Could Make Better Games?

Do you think that there there any technological advances that have yet to happen that could enable better video games? Or do we already have the capability to make games as good as they can be?

That’s hard to predict. I’d guess that we’ve hit the point of diminishing returns. There will certainly be improvements in computing technology in the future, which will enable more complex calculations, for graphics, AI, physics, etc, but as to whether these will enable better games or simply different games, it’s hard to say. We’ve hit on a standardized style of controller design that I don’t think there can be significant improvements in.

We might invent new technologies for input, like improved motion controls, or mind control, which will solve issues like 3-axis movement (current input devices can only operate on 2 axes) or allow us to operate more than 2 interfaces simultaneously (buttons + axis control usually).

Video Games currently are very much about controlling whole bodies, rather than gross or fine motor control and this is reflective of our means of input. Motion control games with 6DOF inputs have allowed us to explore gross motor control slightly more in video games, but we’ve had a hard time making games that match this control scheme, in part because of the lack of force feedback.

Games have suffered from graphical limitations in the past, which limited which types of gameplay were feasible for an assortment of reasons, but most of these limitations have been lifted in more recent times. I think almost any fundamental unit of gameplay that can be achieved now has been achievable for the past decade. Graphics do allow for more objects to be visually represented than before, which is a big deal for MMOs and RTS. Future improvements in graphics that enable new gameplay technologies will probably manifest themselves as improvements in softbody physics, fluid simulations, or other dynamic effects that were previously difficult to simulate. It seems unlikely to me that these will become widespread however.

Newer networking technologies, and the rollout of higher internet speeds could potentially have a massive influence on MMOs in the future, a genre that is harshly limited by bandwidth and response times.

But largely in terms of developing better video games, I think we’ve largely hit the point like traditional 2d animation did where we have all the tools we need to deliver high quality products, and significant technological innovation isn’t really going to fundamentally change the nature of the process anymore. It comes down to using the tools we have better rather than developing new tools. Of course, I’m not psychic, I don’t know what’s coming next, but we’ve seen a mostly steady state since generation 6, and almost completely since generation 7. Gen 8, which we’re on now, isn’t really that different from Gen 7, and it seems unlikely that Gen 9 will be either. There’s a lot more that can be done with the tools we have, but in terms of technological innovation, I don’t see nearly as strong a potential for a revolution that will change the entire industry.

Fighting on Pad vs Stick

What kind of fight stick/gamepad do you use for fighting games, and why?

I use a Dualshock 3. I mostly use it because it’s what I used when I started playing fighting games (although technically I first used the Wii Classic controller for SF2 Hyper Fighting and Guilty Gear XXAC), and because I really like having a good Dpad in the top position.

Pad means my inputs are mostly silent, and I have very good control over my movement, but not quite as precise control over my directional inputs as I might get on stick. It also means that I have difficulty pressing square + triangle or cross + circle and need to bind macro inputs for those usually, as well as R1 + R2.

If I had a consistent way to, I’d ideally use a dualshock 2 controller. Unfortunately my converter is spotty, so that’s not really a viable option (which sucks, because the converter I bought is known for being reliable too). Probably the biggest downside of the DS3 controller is the analog trigger. It doesn’t feel very comfortable, and it’s difficult to know what distance actually triggers the actuation, but I have gotten used to it over time. The DS4 trigger is much more comfortable, but the actuation distance is further, so it throws off all my timings when I use a DS4 pad. Dualshock 2 has a digital button for the trigger, which is highly preferable. Both the DS2 and DS3 have very rough Dpads, and I have actually rubbed my thumbs raw multiple times playing fighting games. I eventually sanded down the dpads to get a smoother surface to play on. The DS4 starts out smooth, avoiding this problem.

I own a Hori RAP4 fight stick. I tried using it, but didn’t stick it out or put the time in to really get good with it. I hold the stick with the same grip as Daigo, from the bottom, between the ring and pinky fingers. I have a lot of difficulty canceling into 2QCF supers or shoryukens on stick. I still hang onto the stick to let other people use it, and I’ll probably learn to play on it so I can play on arcade machines someday.

I think 360 inputs are probably the hardest input to perform on pad relative to stick. Sticks are much better at half circle motions in general, where I think pads are better for DPs and QCFs. Some older games (like 3rd strike, and older versions of guilty gear and blazblue) only count the cardinal directions as valid for their 360 and half circle motions, so if you miss the down input in a half circle back, (6314 instead of 63214), it won’t read your input. (624 does count as valid, it does not care about the 3 or 1 directional inputs) This can happen by doing the motion too fast on pad, where on stick it’s basically impossible. I learned how to do 360s in 3rd strike only after I went into training with input display and realized this was the case. My technique for doing them on pad is to do a half circle back, then tap up and punch, greatly improving my consistency. Modern games like SF4 and SFV let me churn the butter without needing to worry about inputting the cardinal directions accurately.